Dave CROCKER wrote:
John Levine wrote:
You're both right. The open issue in ADSP at this point is whether to
overload i= to attempt to add per-address reputation, or strip out the
cruft so it really does say this domain signs all mail.
wouldn't that mean defining ADSP in terms of d=, rather than i=?
it would also remove the confusion about granularity.
That's only one way to remove some of the overloading of i=; it doesn't
follow that ADSP needs to change. The implication here seems to be that
it's ADSP that is doing the overloading. There has been a lot of (in
many cases speculative) discussion about how the i= value could be used,
but the ADSP specification is the only place where an application for i=
is actually described (except possibly in Doug Otis's drafts).
-Jim
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html